Position: Intelligent Coding Systems Should Write Programs with Justifications

TL;DR

This paper argues that intelligent coding systems should generate not only code but also clear justifications to improve trust and usability, especially for non-expert users, by exploring neuro-symbolic approaches for explanation generation.

Contribution

It introduces the concept of justification properties for AI coding systems and advocates for neuro-symbolic methods to produce transparent, consistent explanations.

Findings

Existing methods like formal verification and static analysis have limitations.

Neuro-symbolic approaches can enhance justification quality during training.

Automated consistency checks can improve trustworthiness of AI-generated code.

Abstract

Intelligent coding systems are transforming software development by enabling users to specify code behavior in natural language. However, the opaque decision-making of AI-driven coders raises trust and usability concerns, particularly for non-expert users who cannot inspect low-level implementations. We argue that these systems should not only generate code but also produce clear, consistent justifications that bridge model reasoning and user understanding. To this end, we identify two critical justification properties-cognitive alignment and semantic faithfulness-and highlight the limitations of existing methods, including formal verification, static analysis, and post-hoc explainability. We advocate exploring neuro-symbolic approaches for justification generation, where symbolic constraints guide model behavior during training and program semantics are enriched through neural representations, enabling automated consistency checks at inference time.